DNA arrays provide a broad snapshot of the state of the cell by measuring the expression levels of thousands of genes simultaneously. Visualization can provide effective tools to summarize and interpret data sets, describe the contents, and expose features in time series data. Array-derived gene expression data sets present formidable analysis and visualization challenges because of their dimensionality. The Fourier harmonic projections (FHPs) were used to map multi-dimensional gene expresion data to two dimensions in an implementation called VizStruct. These projections have useful properties that preserve the certain key characteristics of a variety of data sets and offer strong cluster delineation capabilities.
The objective of VizStruct is to perform a visual exploration of gene expression data sets through two processes: (1) Explorative Visualization Process which focuses on sample space and presents an interactive visualization clustering approach to classify different samples based on variations of gene expressions, and (2) Confirmative Visualization Process which concentrates on gene space and applies Fourier harmonic projections to investigate the structure of the data set and evaluate the effectiveness of the class assignment done by a variety of clustering algorithms.
